07:10 <emilbayes> SeanTAllen: Thank you for your patience

07:10 <emilbayes> SeanTAllen: ByteSeq is exactly the type i want

07:23 <emilbayes> SeanTAllen: Is there a general pattern for taking in a ByteSeq and filling it with data? Specifically which ref caps do I use? ref? iso? I tried to gist my idea here: https://is.gd/u73SQ2

08:25 <Candle> SeanTAllen: Yeah, I agree. I don't see any fixed point or arbitrary precision (other than using an integer and arbitrarally fixing a decimal point) I'm not doing financial computations, so I'm happy with the inaccuracy. There may be a PR against the TestHelper to add some better support.

12:37 <SeanTAllen> emilbayes: check out how the read buffer works in TCPConnection. That might be what you are looking for.

12:37 <emilbayes> SeanTAllen: Ok, will do!

12:37 <emilbayes> SeanTAllen: Feel like I tried to read a lot of code, but didn't check that one :)

12:38 <SeanTAllen> Candle: I'm assuming that that error is a formatting issue and the numbers aren't equal but they are "equal" at that lesser level of precision that is being displayed.

12:40 <Candle> SeanTAllen: That's a safe assumption. The PR will probably add TestHelper.assert_between or *something* similar.

12:41 <tokenrove> usually you want to compare within an epsilon, where the epsilon may be related to the ULP

12:41 <Candle> Given that the actual number was comming from the parsed json (thanks for your help there!) and the expected from the unit test; I was surprised that it failed.

12:43 <Candle> tokenrove: Yes, like junit assertEquals(double expected, double actual, double epsilon).

12:44 <tokenrove> Candle: just in case you haven't seen it, bruce dawson's article about this is good reading: https://randomascii.wordpress.com/2012/02/25/comparing-floating-point-numbers-2012-edition/

13:49 <TwoNotes> Apparently something changed in the syntax for an array literal.

13:53 <Candle> TwoNotes: s/,/;/g

13:53 <TwoNotes> That is what I thought, but it does not work when the array literal spans more than one line

13:53 <Candle> TwoNotes: ah, yes, then the ; is omitted.

13:54 <TwoNotes> So the REAL separate is just a space - it is a series of expressions within square brackets. The ';' is there to separate expressions on one line?

13:54 <Candle> afaik, one element per line and no ;. any that are on the same line are separated by ;.

13:55 <TwoNotes> ok

13:57 <Candle> tokenrove: Given his conclusions; providing a 3-arg assert is the easiest; delegate that decision to the 'user'.